Journal of Clinical Laboratory Analysis 7994-400 (1993)
Protein 1 : Its Purification and Application in Clinical Medicine
Yoshihisa Itoh,’ Shuichi Ishii,’ Ryuta Okutani,’ Yasushi Asano,2 and Tadashi Kawai’
Department of ‘Clinical Pathology and *Nephrology, Jichi Medical School, Tochigi-Ken, Japan
Protein 1 ( P l ) is a low-molecular-weight
protein recently isolated from the urine of pa-
tients with chronic renal failure. Its molecu-
lar weight is 14 kDa on sodium dodecyl
sulfate polyacrylamide gel electrophoresis
and pl 4.7 on isoelectric focusing. We puri-
fied this protein, characterized its physico-
chemical properties, and analyzed its amino
acid sequences to show that it is probably
identical to human lung Clara cell 10 kDa pro-
tein. Its monoclonal antibody was prepared,
and a reliable enzyme-linked immunosorbent
assay employing the sandwich method was
developed and used to investigate distribu-
tion and variation in concentration of P i in
Key words: chronic renal failure, genital tissue, human Clara cell 10 kDa protein, protein 1,
renal tubular dysfunction
INTRODUCTION
Protein 1 (Pl), also called urine protein 1 (UPl), an
electrophoretically a2-migrating nonglycoprotein with a
molecular weight of 14 kDa and PI 4.7, was initially isolated
from pathologic urine and characterized by Jackson et al. (1)
(Tablc 1). A partial amino acid sequence showed this protein
to have 55% homology with rabbit uteroglobin (1). Clinical
studies were subsequently conducted by Bernard et a]. (4),
who used a latex immunoassay to demonstrate the stability
of PI in urine and marked sex-related differences in its uri-
nary concentration ( 2 , 3 ) .Variation in urinary levels of P1 were
also investigated in patients with cadmium poisoning (4).
We have purified low molecular weight protcins (LMWPsj
from urine and developed new systems to assay them and de-
termine their clinical significance (5,6). When sodium dodecyl
sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was
used to analyze concentrated urine from patients with renal
tubular disorders, a faint but discrete band frequently appeared
at a molecular weight of 15 kDa, close to the lysozyme band.
Our preliminary study used urine fractions obtained in the
purification of LMWP to determine the presence or absence
of this protein. The protein-rich fraction thus obtained almost
always occurred in the fractions that were shown to contain
retinol binding protein (RBP) on gel-filtration, a finding con-
firmed by SDS-PAGE. Ouchterlony immunodiffusion was used
to investigate immunoreactivity between the concentrated frac-
tions and various commercially available antisera. A strong
precipitin line was observed with rabbit anti-P1 antibody
0 1993 Wiley-Liss, Inc.
various body fluids under an array of physio-
logic and pathologic conditions. Clinical stud-
ies indicated that, as is the case with other
proteins of low molecular weight, the main
catabolic site of P1 of plasma origin is the
kidney: P1 is filtered by the renal glomeruli
and reabsorbed by the renal tubules. Unab-
sorbed P1 is thus excreted into the urine. This
protein is also synthesized in the genital tis-
sues of males, however, from which it is also
excreted into the urine. Clinical data obtained
in our study of this protein is summarized
here, and an attempt is made to determine
the potential value of this protein in labora-
tory medicine. o 1 9 9 3 ~ i 1 e y - ~ i sInc.
s,
(DAKO, Copenhagen), with which the precipitin line pro-
duced by RBP did not cross-react (unpublished data). Using
previously established procedures (5,6), we started out study
of P1 by isolating this protein from pathologic urine, aiming
to establish its clinical significance as a new serum and uri-
nary marker for diseases.
ISOLATION OF PROTEIN 1 FROM URINE
RBP deprivation was a critical problem in an initial purifi-
cation study, because its physicochemical properties of RBP
(M.W. 21 kDa, PI 4.7) were quite similar to those of P1.
Furthermore, the concentration of RBP in the urine that served
as a purification source was 20 times higher than that of P1
(3,7). The authors therefore adopted anti-RBP coupled af-
finity chromatography and used it in combination with am-
monium sulfate precipitation, concanavalin-A affinity, gel
filtration, and anion-exchange column chromatography. RBP
was successfully removed and highly purified P1 was obtained
(3,7). Recovery was greatly improved after we further sim-
plified the purification step by first using ammonium sulfate
precipitation followed by anion-exchange chromatography,
anti-P1 antibody coupled affinity chmmatogmphy, and reversed-
phase HPLC, in this order, to prepare mouse monoclonal an-
tibody (Fig. 1): 41.6% in the initial purification; 60.2% in
Received June 8, 1993; accepted June 10, 1993.
Addreqs reprint requests to Dr. Y. Itoh. Department of Clinical Pathology,
Jichi Medical School, Minami-Kawachi-Machi, Tochigi-Ken, Japan.
 Protein 1 395
TABLE 1. Essentials of Protein 1
Molecular weight 14 kDa on SDS-PAGE
20 kDa on gel-Ciltration
Pi 4.7
Electrophoretic mobility a2
Structure Non-glycoprntein
Identical with CClO
Homodimer of 70 ammo acids
55% homology with rabbit UG"
Synthesis ClWd cell
Genital tissues
Function PLAZ inhibitorb

"UG: Uteroglohin
'Data from Clara cell 10 kDa protein

the simplified purification. The isolated substance was iden-

tified as P1, because its physicochemical properties and struc-
ture were the same as those previously reported (1): molecular
weight of the purified substance was 14 kDa on SDS-PAGE.
Upon reduction by 2-mercaptoethanol, a single band appeared
at a molecular weight of 7 kDa, showing this protein to be a
homodimer of a 7 kDa subunit through disulfide bridge link-
age (Fig. 2). Its isoelectric point was PI 4.7 on isoelectric
focusing. The N-terminal amino acid sequence of this pro-
tein was determined up to 53 amino acids, identical to that
initially reported by Jackson et al. (1).
Information found in a protein database showed that the
amino acid sequence of PI we had determined to be identi-
cal to that of human lung Clara cell 10 kilodalton protein
(CClO). CClO produced in noncilliated cells in the lung con-
sisted of a dimer of 70 identical amino acids and had a mo-
lecular weight of 10 kDa; PI. 4.8 (8-10). Slight differences
in molecular weight between the two are accounted for by
the difference in buffer used in the SDS-PAGE. The physico-
Fig. 2. Purified protein 1 on SDS-PAGE and Western blots. Marker pro-
tein (lanes 1 . 6 ) , purified prolein I under nonreducing conditions stained
with Amido Black 10B (lane 2), and indirect immunostaining with a
monoclonal antibody (lane 3). Purified protein 1 under reducing condition
stained with Amido Black 1OB and indirect immunostaining with same an-
tibody (lanes 4,5) (7). (With the permivsion of Elsevier, Amsterdam.)
chemical properties and structure of the isolated substance
suggest that P1 is identical with CC10. Bernard et al. (1 1 )
have also recently reported the identity of PI and CC10. This
discovery was especially important in that CC 10 functions
as a phospholipase A2 (PLA2) inhibitor regulating initia-
tion of an arachidonic-acid cascade.

AN ENZYME-LINKED IMMUNOSORBENT ASSAY

Procedure for the purification of P1 from urine
Bernard et al. (2) established a precise, automated latex
Pathologic urine immunoassay using an antigen and antibody from DAKO.
1 60% saturated (NH,),SO,
Precision and recovery were both good. Thc sensitivity limit
Supernatants
1 90% saturated (NH,),SO,
was 0.5 Fgil. The authors have also established a new cnzyme-
Precipitates (Urine concentrate) linked immunosorbent assay employing the sandwich method
1 Diaiyzed against lOmM Tris-HCI (pH7.4) (ELISA) for this protein (Fig. 3 ) (12). The standard used in
Sephadex A-25 anion-exchange chromatography this ELISA was purified P1, and its concentration was deter-
I Eluted with a linear gradient of 0-0.3M NaCl in
lOmM Tris-HCI (pH7.4)
Anti-PI immunoaffinity chromatography
mined by Coomassie protein assay reagent (Pierce, IL) using
bovine serum albumin as a standard. The molecular extinc-
tion coefficient of the purified substance at 1% (wiv) was

I
Eluted with 0.2M Glycine-HCI (pH2.0)
Neutralized by dialysis against 1OmM Tris-HCI 6.0 at 280 nm. Each well of a plastic plate was coated with
(pH7.4) our monoclonal antibody. A rabbit antibody (DAKO) was used
Reversed-phase HPLC to react P I , which was bound with the monoclonal antibody.
I
Purified P1
Eluted witha lineargradient of 0-100%CH3CN i n
O.OS%TFA
Horseradish peroxidase conjugated-goat antibody against rab-
bit antibody (American Qualex, CA) was then added. Last,
(Recovery: 60.2% against urine concentrate) en7ymatic activity was measured at 492 nm using the
o-phenylenediamine/H202system. The present assay had ex-
Fig. 1. Purification proccdurc from pathologic urine cellent sensitivity and was able to detect 5 pgiml of P1: the
396 ltoh et al.

coating mAb on a plate

T1 r pH 8. P1 was quite stable in both normal and pathologic urine.
When pooled pathologic urine from patients with renal fail-
ure was stored at pH 4 to pH 8, both at 25°C and 4"C, pro-
tcin value remained unchanged for 4 days (Fig. 4a,b). and
the same was true for pooled normal urine (data not shown).
It is important to note that nonglycoprotein was also quite
stable in acid urine. (j2-Microglobulin (Pz-m), another
1 nonglycoprotein with a molecular weight of 11 kDa, is
incubating at RT for 9 0 mln.
quite unstable, losing its antigenicity within 1 day when
held at 37°C in acid urine (14). The structure of proteins
that function as substrates for urinary proteases warrants
adding 100111
polyclonal Ab
Of
111 further study.

1
Incubating at RT for 90 min.

=
a -1
adding 1OOpl of
.
E
0
E

peroxidase-conjugated 5 300- PH5

,.~/:!.. lio1l
anti-rabbit IgG goat polyclonal Ab
-3 Q)

0
incubating at RT for 40 min. > 200

1 h
~

pH4-8,4 "r:
adding 100pl of
&d.&
(o-phenyIenediamine/H,O,)
- 0
1 0 1 2 3 4 5
reacting at RT for 6 min.
1
of 1M H2S0,
stopping reactlon wlth 1 5 0 ~ 1
1
measured at 492nm

Fig. 3. An ELISA procedure. RT: room temperature

range was SO pg/ml to 5,000 pgiml. Precision, recovery, and

specificity were good. The latex assay used here indicated a
higher serum P1 concentration in normal adults than did the
present ELISA (S,7). Both indicated similar physicochemi-
cal properties of the purified substance. Further study will
be necessary to settle this problem by exchanging purified -a
Q
forms of each substance. Differences in the principles on which >
200 -
the assays were based, as well as calibrators, antibodies. and
h
the buffers used may explain the differences observed.

STABILITY IN NORMAL AND PATHOLOGIC URINE 100 - pH4-8,25 't:

Protein stability is of critical importance in urinalysis: with-
out it, protein value is apt to be underestimated. Although
the precise mechanisms are unknown, the structure-related 0 1 2 3 4 5
characteristics of proteins and protease activity in urine are
thought to be the most important factors determining stabil- Day
ity of urinary proteins. Stability of PI in normal and patho- Fig. 4. Stability of protein 1 in pathologic urine. (a) pH 4-8, 4°C; (b) pH
logic urine was investigated at 25°C and 4"C, from pH 4 to 4-x. 25°C.
 Protein 1 397
DISTRIBUTION OF PROTEIN 1 IN VARIOUS greater individual variation, which later proved to reflect de-
BODY FLUIDS gree of excretion of PI from genital tissues into the urine from
genital tissues as well as the sequential fraction from which
A previously developed ELISA was used to determine the
the sample was obtained.
distribution of P1 in various body fluids under normal and
Urinary excretion was elevated in cases of normal preg-
pathologic conditions (Table 2). Marked sex-associated dif-
nancy (1 1). Increase in urinary P I values appeared during
ferences were noted in adult urine, but not serum. P I levels
the third trimester and again just before the initiation of de-
were highest in bronchioalveolar lavage fluid (BALF) obtained
livery. This may be partially explained by the increased per-
from patients with asthma, sarcoidosis, and pulmonary fi-
meability of the glomeruli, physiologic changes in renal tubular
brosis. P I , which is identical to CC10, is thought to be ac-
function, and increase in pre-renal overload. Contamination
tively synthesized and secreted into the bronchiolar lumen in
from genital tissues and amniotic fluid may also contribute
association with these pulmonary diseases. P1 may have an
to its elevation in urine prior to delivery. No significant ele-
anti-inflammatory effect, probably by acting as a PLA2 in-
vation in serum values was seen throughout pregnancy. At
hibitor. P I also occurs in high concentrations in some gastric
termination of pregnancy, parturition is started by the stimu-
Iluids. When a gastric stimulation test was performed by in-
lationof prostaglandinF2a (PGF2a), estradiol, oxytocin, and
jection of gastrin, the values observed in the cases of gastric
placental decrease in progesterone. Elevation of P1 concen-
diseases investigated were highest in some instances of chronic
tration in the urine may be associated with maintaining preg-
gastritis in which concentration of P1 and pH level were both
nancy by inhibiting contraction of the uterus through the
high. Thus P1 may have been present in the original gastric
regulation of PGF2a activity by inhibiting PLA2 activity.
juice and have been digested by the action o f pepsin in an
Urinary excretion of PI was elevated after certain forms of
acid environment. It is unknown whether P1 is locally pro-
strenuous physical exercise. In males, this increase may re-
duced in the stomach or not. PI can be of lung origin simply
sult from activities like bicycling, in which the genital tis-
as a result of the swallowing of P1-rich sputa. Further, this
sues are compressed.
protein also occurs in minute quantities in other body fluids.
SYNTHESIS OF PROTEIN 1 BY GENITAL TISSUES
PHYSIOLOGICVARIATION
An earlier study pointed out that urinary P1 concentration
Serum P I level is constant and shows no sex-associated
was higher in males than in females (2,3). Although this dif-
differences from the third through the seventh decades. This
ference was not found in elementary school pupils ranging
stands in sharp contrast with sex-associated differences in lev-
from 7 to 9 years of age, significant differences appeared
els of P1 in the urine: age and sex-associated variation of PI
among junior high school students 13- 15 years old. In nor-
was found in spot urine and was higher in male than that in
mal adults of reproductive age. PI values are 10-20 times
female subjects (Table 2). Pre-renal load on the kidney prob-
higher in males than in females (3). Among sequentially col-
ably varies little with sex and may thus have little effect on
lected urine samples, the highest concentration was found in
sex-associated differences in urinary P1 values.
the first sequential fractions. P1 concentration in the remain-
No diurnal variation in serum P1 concentration was found
ing fractions was s o low that it approached that of urine from
in healthy individuals upon whom n o restrictions on daily
females, which remained constant in all fractions (Fig. 5).
routine had been placed. When whole urine was collected
This result strongly suggested that the genital tissues were
over the course of a day, its values in males showed much
the origin of P1 and that P1 was present in high concentra-
tion in the semen (1 1,13). On Western blots, molecular weight
of P1 in the semen was 14 kDa but did not have an additional
TABLE 2. P1 Values in Various Body Fluids band in the low or high molecular weight position (14). P1
Samples No Mean & lSD(pg/l) Range(pgi1) was also found in seminal vesicle fluids, though no P I was
Umbilical cord serum contained in the spermatozoa. The use of immunochemical
Male 12 9.9 i 3.2 5.6 ~ 16.3 methods to search for synthetic sites suggested that the geni-
Female 12 9.6 ? 3.3 5.5 ~ 13.4 tal tissues are one source of P I in males (1 3). A recent im-
Serum of normal adults
Male 9 11.824.7 3.0 - 19.1
munohistochemical study has shown localization of this protein
Female 9 13.4 = 4.0 8.3 - 19.1 in the epithelium of seminal vesicle, ejaculatory duct, and
Spot urine in normal adults possibly the prostate (see Fig. 7). Sex-associated difference
Male 18 14.2 * 13.9 1.7 - 42.7 in P1 concentration in adult urine is clearly associated with
Female 18 1.0 5 1.2 0.2 - 4.2
Gastric juices 10 185.2 -C 3665.7 0 - 1221.8
time of voiding: P1 secreted from male genital tissues was
Rronchioalvcolar lavage fluids 16 1222.2 2 630.1 147.2 2677.3
~
mixed with structurally identical PI of plasma origin stored
Cerebrospinal fluids 10 0.4 5 0.8 0 - 2.1 in the bladder.
Svnovial fluids 4 4.6 * 0.4 4.2 - 5.2 Extraordinarily elevated urinary concentration of P1 was
398 ltoh et al.
- No1
v
0
a
200
- No7
-e
> sults indicate that P1 is catabolized in the kidney in the
100
0
0 1 2 3 4 5
Order of Urine Colleclion
Fig. 5. PI values of four sequentially collected urine samples with seven
normal males.
observed in patients with prostatic cancer, and this elevation
was especially pronounced in the first sequential urine frac-
tion. Use of this protein as a tumor marker may warrant fur-
ther investigation.
KIDNEY FUNCTION AND PROTEIN 1
Urinary excretion of PI was elevated in patients with cad-
mium nephrotoxicity (4). Although urinary concentration of
PI varied but did not differ significantly from that of con-
trols, increased values were slightly less frequent in the con-
trol group. Excretion of this protein correlated closely with
concentration of urinary cadmium. Concentration of albumin
and excretion of transferin were greater in the earlier stage
of nephrotoxicity than was concentration of P1, indicating
that impaired glomerular dysfunction in combination with tu-
bular dysfunction may be ashociated with elevated concen-
tration of urinary P1 (4). This suggests that handling of PI at
the renal tubule may be possible, though not decisive. Early
preliminary observations showed P1 excretion to be elevated
in renal glomerulotubular dysfunction. This study began by
evaluating the renal glomeruli by measuring variation in se-
rum P1 values in renal diseases. No disease-specific changes
were observed; howeyer, serum concentration of PI was
strongly associated with serum creatinine (Cr) concentration.
A good correlation was also observed with Cr clearance,
clearly showing that this protein was filtered through the glo-
merular basement membrane like other LMWPs. Reabsorp-
tion of P1 at the renal tubules was directly demonstrated by
the L-arginine infusion test (Fig. 6) (14), in which L-arginine
monohydrochloride was infused in six normal adults and PI
value was measured every 5-10 minutes along with other
LMWPs, N-acetyl-P-D-glucosaminidase (NAG), and albu-
min. As excretion of L-arginine increases, thus blocking re-
absorption of proteins at the proximal tubules, elevation in
P 1 excretion closely paralleled elevation in a I -microglobulin
(cu,-m), Pz-m, and other LMWPs, whereas concentrations of
albumin and NAG were almost within normal limits.
Decrease in L-arginine excretion resulted in a closely
parallel decrease in all LMWPs. All of these clinical re-
same way as other LMWPs.
Actual handling of P1 by the kidney was re-evaluated us-
ing 24-hour-stored urine taken in three sequential fractions
from each individual. PI present in the first of the three frac-
tions is thought to be of both plasma and genital-tissue ori-
gin, whereas that in the last probably comes only from plasma
stored in the bladder. Twenty-four-hour excretion of P1 from
the two different sources was thus calculated. Urinary excrc-
tion and the ratio between PI derived from the plasma and
that derived from the genital tissues were constant when the
urine samples were collected over 2 consecutive days. Cal-
culated concentration of PI derived from the plasma alone
was 4.1 -+ 1.3 kg/day; that in the remaining fractions, which
originated in the genital tissues, was 5.8 2 3.1 pg/day. The
ratio between the two fractions was 0.98 ? 0.82 and ranged
from 0.43 to 2.06. Actual fractional clearance, defined as
P1 clearance divided by Cr clearance. was 0.0020 f- 0.0008
inmales, almost the sameas that infemales, 0.0013 & 0.0019.
This actual value is also very close to those for other LMWPs.
We thus concluded that P1 stored in bladder is mainly of
plasma origin and is filtered through the kidney. The rest is
derived from genital tissues. Urine from the second and third
of the three fractions thus contains P1 that was mainly of
plasma origin. P1 can thus be used as a marker for evalua-
tion of renal function, even in males, as long as the urine is
collected carefully. Restriction of sexual activity and physi-
cal exercises would also be necessary, however. For women
and prepubescent children, urinary P1 is entirely of plasma
origin and is, like other LMWPs, therefore useful without
restriction in sample collection (13).
A recent study using urine from females has shown uri-
nary P1 concentration to be elevated in patients with diabetic
nephropathy and in very good accordance with that of albu-
min. Dynamic elevation over 1.000 ~ g iofl urinary excretion,
which closely paralleled that of other LMWPs (unpublished
data), was also demonstrated in pediatric patients with renal
tubular dysfunction. Structurally identical P1 from two
different sources is present in the urine of males. We
are now using sequentially collected urine fractions to
re-evaluate renal function and to detect those physio-
logic and pathologic changes in genital tissue that may
be relevant to functional aspects of P1 as an inhibitor
of PLA2, which restricts the enzyme that initiates the
arachidonic acid cascade.
 Protein 1 399

T h e (minutes)

Fig. 6. Elevated excretion of urinary protein I in good padlel with Lu,-microglobulin (cl,-ni) and P2-microglobulin(P2-m) by L-Arginine infusion (14).
(With the permission of S . Karger, Basel.)

ACKNOWLEDGMENTS
This work was supported by Grant-in-Aid from the Minis-
try of Education (Project No. 02454490, 04671439) and a
Research Grant from the Specific Diseases Division from Pro-
gressive Renal Disorders, Ministry of Health and Welfare,
Japan. The authors are grateful to Miss Tomomi Kebukawa
for her excellent technical assistance.

REFERENCES
1. Jackson PJ, Turner R, Keen JN. Brooksbank RA. Cooper EH: Purifica-
tion and partial amino acid sequence of human urine protein 1. J
Chromutugr 452:359-367. 1988.
2. Bernard AM, Lauwerys RR, Noel A, Vandeleene B, Bernard RR: A
sex-dependent marker of tubular or glomerular dysfunction. Clin Chem
35:2141-2142, 1989.
3. Itoh Y. Ishii S, Okutani R, & Kawai T: Purification of urine protein 1
(UP 1) from urine: An initial study toward clinical application. Prog-
ress in Clinical Biochemist?; 307 -3 11, 1992.
4. Bernard AM, Roels H, Cardenas A & Lauwerys R: Assessment ofuri-
nary protein I and transferrin as early markers of cadmium nephrotox-
icity. BrJIndMed 47:559-565. 1990.
5 . Itoh Y, Kawai T: Human a I-microglobulin: Its measurement and clini-
cal significance. J Clin Lab Aizul4:376-384, 1990.
6. Itoh Y, Enomoto H. Kawai T: A latex agglutination test for p2-micro-
globulin in urine. ContrNephrol68:172-178, 1988.
7 . Okutani R , Itoh Y, Hirata H, Kasahara T, Mukaida N, Kawai T: Simple
and high-yield purification of urine protein I using iinmunoaffinity chro-
matogrdphy: evidence for the identity of urine protein I and human Clara
cell 10-kilodalton protein. J Chrornatogr 577:25-35, 1992.
8. Singh G, Singh J, Katyal SL, Brown WE, Kramps JA, Paradis IL,
Dauber JH. Macpherson TA, Squeglia N: Identification, cellular local-
ization, isolation, and characterization or human Clara cell-specific 10
KD protein. J Histochetn Cytochern 36:73-80, 1988.
9. Singh G, Katyal SL, Brown WE. Philips S ? Kennedy AL, Anthony J.
Squeglia N: Amino-acid and cDNA nuclcotide sequences of human Clara
Fig. 7. Immunochcmical detection of protein I in the seminal vesicle. cell 10 kDa protein. Biuchim Biuphys Acm 950:329-337, 1988.
400 ltoh et al.
1 0 . Singh G. Katyal SL. Brown WE, Kennedy AL. Singh U, Wong-
Chong M-L: Clara cell 10 kDa protein (CCIO): Comparison of
structure and function to uteroglobin. Biuchim Bzophys A m 1039:
348-355. 1990.
1 1 . Bernard A, Roels H , Lauwerys R . Witters R . Gielens C, Soumil-
lion A , Damme JV. Ley MD: Human urinary protein 1 : evidence
for identity with the Clara cell protein and occurrence in respir-
atory tract and urogenital secretions. Ctin Chim Acfu 207:239-
249, 1992.
12. Ishii S, Itoh Y, Okutani R. Asano Y, Kawai T: Developnient of an
enzyme-linked immunosorbent assay for protein 1. Contr Nephrol
101:71-77. 1993.
13. Asano Y. Itoh Y, Kawai T: A study on protein 1 in urine: In An Annucd
Report on Progressive Renal Disease. Kiyoshi Kurokawa, ed. Ministry
of Health and Welfare, Tokyo, 1992. pp 75-78.
14. ltoh Y, Muto S. Okutani R . lshii S, Asano Y. Kawai T: Increased uri-
nary excretion o f urine protein 1 induced by L-arginine monohydro-
chloride. Nephron 62467-468, 1992.